A major goal of breast cancer research is the identification of cancer stem cells (CSCs), and characterization of their functional properties wit a view to targeted therapy. Based on compelling recent data from the small intestine and mammary gland, we will test the hypotheses (a) that expression of the R-spondin receptor Lgr5 defines a population of mammary tumor stem cells with highly efficient tumor- initiating capacity; (b) that, following radiotherapy in a model system, the dose-response profile of Lgr5+ cell survival predicts that of tumor elimination; and (c) that Lgr5+ cells act as stem cells for mammary tumor regeneration after non-ablative radiation. The research will exploit experimentally tractable mouse models of breast cancer and a state-of-the-art small animal microirradiator that delivers uniform doses of irradiation to tumors with negligible damage to surrounding tissue. To facilitate identification and isolation of Lgr5+ tumor cells, transgenic mic with a mammary-targeted oncogene will be crossed with Lgr5- reporter strains. Lgr5+ cells will be sorted from the resulting mammary tumors by flow cytometry and their tumor-initiating capacity measured by injecting recipient animals with limiting cell dilutions. Lgr5+ cells will be compared with Lgr-, and with mock-sorted tumor epithelial cells, to determine whether the bulk of tumor-initiating capacity is associated with Lgr5+ cells. Since Lgr5 is an accessory receptor for the Wnt/?-catenin pathway, the protein may serve not only as a stem cell marker but also as a functional determinant of CSC properties. This notion will be tested in tumorsphere assays of CSC self-renewal in vitro. Further in vivo experiments will test the hypothesis that Lgr5+ cells are CSCs which govern the tumor's long-term response to radiation therapy. Using a CT image-guided small animal microirradiator to deliver single high doses to mouse mammary tumors, the dose-response profile of Lgr5+ cell elimination following irradiation will be studied in the contex of local tumor control. This will test the hypothesis in vivo that extinction of the tumor depends on elimination of Lgr5+ cells. Further evidence of the therapeutic relevance of these cells will be determined from lineage tracing experiments asking whether regrowth of tumors following non-ablative irradiation occurs from Lgr5+ CSCs. If the expected results are obtained, they will identify Lgr5+ cells as CSCs that are critical for both initiation of tumorigenesis and for tumor survival/growth following radiation treatment. Thus we expect to identify a therapeutically relevant key population of CSCs. Moreover, since Wnt signaling inhibitors constitute a major new category of investigational drugs, these may provide a therapeutic means of targeting Lgr5+ CSCs and rendering them susceptible to elimination.